4 ms·

"In the case of fdroid, apps that violate open source licenses are not allowed" ...on the main repository. AFAIK, there's nothing stopping Google or anyone els

"In the case of fdroid, apps that violate open source licenses are not allowed" ...on the main repository. AFAIK, there's nothing stopping Google or anyone else from setting up their own F-Droid repository to distribute apps with proprietary code. The normal F-Droid app should be able to use a repository like that just fine. EDIT: Addressing the "PS" that was added... > Google advocates always make the argument that endusers "can just root their phones and install the APKs anyways" which is similar to f-droid with an external repository. Most non-technical endusers simply won't do that. Android skirts around the criticisms fielded towards iOS by technically allowing users to install and distribute third-party apps. The real problem with Android is that the default distribution platform (Google Play Store) is a walled-garden, proprietary app with such a massively disproportionate market share that most users don't even realize there are alternatives. And Google ensures their store will always be the default because they hold their proprietary Google Play Services for ransom. And Google Play Services is so valuable because it provides many convenient features and functions, including some which used to be part of the operating system itself.

cookiengineer 5y ago
I totally agree with your points there. But I think that the main issues of Android (or AOSP) are even a level deeper than just the Play Services. There are lots of initiatives that try to create a free ecosystem for themselves (Lineage, /e/, Carbon, et al), with their own stores and sources for Apps. Most of them have varying degrees of success, due to gapps counterparts like microG [1] not being able to keep up with what Google's Play Services provide API-wise. It's an absurd amount of features, and a lot of API workflows to consider. Bugs and crashes everywhere down the user experience...but hopefully they're getting slowly to a stable state. Coming back to the real problem: I think it's actually the Vendor deals that Google did. Most of the manufactured devices are almost impossible to flash without reverse engineering skills, and this is intentional. Having to wait more than 3 months to unlock a smartphone's bootloader because the manufacturer doesn't give a damn about you is just one of many examples; setting aside that most of the unlock procedures are meant to be understandable by developers-only. I think that in order to "really free Android" the creation, flashing, updating of ROMs has to be standardized in a more homogenic way (partition fatigue, anyone?), because it would allow a graphical and easy-to-use software to be built. That would allow to flash a ROM without e.g. losing all /data and more importantly - be usable by end-users without technical knowledge. In my social circles I'm the guy that flashes LineageOS to their devices, because most of the terminology is so far away from the reality of most users that they have no single clue where to start. The amount of knowledge that is required to flash your device (and be Google-free, even in Apps with e.g. with Appwarden [2]) is absurd and as long as this is the case it will be a niche that's being ignored by politics (and potential regulation laws that would force Google's policies to change). [1] https://lineage.microg.org/ https://lineage.microg.org/ [2] https://gitlab.com/AuroraOSS/AppWarden https://gitlab.com/AuroraOSS/AppWarden
- CivBase 5y ago
  I kind of agree, although I'm not sure it's fair to say that the problem with Android is that you can't easily replace it with another OS. That's not really an Android problem. It's incredible what a smartphone can do given its form factor and a lot of that is thanks to their use of SOCs. I have no experience with OS development for SOCs, but I hear it is much more involved because a new version of the OS must be created for each SOC - specialized to work with the device tree supported by that chip. As I understand, Google doesn't do that work. Manufacturers have to fork Android and implement support for their SOCs on their own, then they have to maintain that fork as new Android releases keep coming. It's no surprise then that manufacturers don't want to invest addition support into other operating systems like LineageOS. There's probably a better way to do things. I'm sure manufacturers could make information more available to OSS communities which would allow them to do the work themselves more quickly and effectively. Like you mentioned, standardization would also go a long way towards making our current smartphone ecosystem more friendly to third-party OSes. But ultimately, none of that is really Android's fault. Even without Google's vendor deals, I doubt the likes of Samsung, Motorola, or any other major smartphone manufacturer would start supporting LineageOS. It's hard enough to even get Linux suppport from desktop/laptop manufacturers. LineageOS is a really amazing project, but I don't think it's the one paving the way for open source operating systems on smartphones. I think most of that work has to come from the hardware side with projects like the PinePhone.

cookiengineer 5y ago

I totally agree with your points there. But I think that the main issues of Android (or AOSP) are even a level deeper than just the Play Services. There are lots of initiatives that try to create a free ecosystem for themselves (Lineage, /e/, Carbon, et al), with their own stores and sources for Apps. Most of them have varying degrees of success, due to gapps counterparts like microG [1] not being able to keep up with what Google's Play Services provide API-wise. It's an absurd amount of features, and a lot of API workflows to consider. Bugs and crashes everywhere down the user experience...but hopefully they're getting slowly to a stable state. Coming back to the real problem: I think it's actually the Vendor deals that Google did. Most of the manufactured devices are almost impossible to flash without reverse engineering skills, and this is intentional. Having to wait more than 3 months to unlock a smartphone's bootloader because the manufacturer doesn't give a damn about you is just one of many examples; setting aside that most of the unlock procedures are meant to be understandable by developers-only. I think that in order to "really free Android" the creation, flashing, updating of ROMs has to be standardized in a more homogenic way (partition fatigue, anyone?), because it would allow a graphical and easy-to-use software to be built. That would allow to flash a ROM without e.g. losing all /data and more importantly - be usable by end-users without technical knowledge. In my social circles I'm the guy that flashes LineageOS to their devices, because most of the terminology is so far away from the reality of most users that they have no single clue where to start. The amount of knowledge that is required to flash your device (and be Google-free, even in Apps with e.g. with Appwarden [2]) is absurd and as long as this is the case it will be a niche that's being ignored by politics (and potential regulation laws that would force Google's policies to change). [1] https://lineage.microg.org/ https://lineage.microg.org/ [2] https://gitlab.com/AuroraOSS/AppWarden https://gitlab.com/AuroraOSS/AppWarden

CivBase 5y ago
I kind of agree, although I'm not sure it's fair to say that the problem with Android is that you can't easily replace it with another OS. That's not really an Android problem. It's incredible what a smartphone can do given its form factor and a lot of that is thanks to their use of SOCs. I have no experience with OS development for SOCs, but I hear it is much more involved because a new version of the OS must be created for each SOC - specialized to work with the device tree supported by that chip. As I understand, Google doesn't do that work. Manufacturers have to fork Android and implement support for their SOCs on their own, then they have to maintain that fork as new Android releases keep coming. It's no surprise then that manufacturers don't want to invest addition support into other operating systems like LineageOS. There's probably a better way to do things. I'm sure manufacturers could make information more available to OSS communities which would allow them to do the work themselves more quickly and effectively. Like you mentioned, standardization would also go a long way towards making our current smartphone ecosystem more friendly to third-party OSes. But ultimately, none of that is really Android's fault. Even without Google's vendor deals, I doubt the likes of Samsung, Motorola, or any other major smartphone manufacturer would start supporting LineageOS. It's hard enough to even get Linux suppport from desktop/laptop manufacturers. LineageOS is a really amazing project, but I don't think it's the one paving the way for open source operating systems on smartphones. I think most of that work has to come from the hardware side with projects like the PinePhone.